SbbHLH85, a bHLH member, modulates resilience to salt stress by regulating root hair growth in sorghum

bHLH family proteins play an important role in plant stress response. However, the molecular mechanism regulating the salt response of bHLH is largely unknown. This study aimed to investigate the function and regulating mechanism of the sweet sorghum SbbHLH85 during salt stress. The results showed that SbbHLH85 was different from its homologs in other species. Also, it was a new atypical bHLH transcription factor and a key gene for root development in sweet sorghum. The overexpression of SbbHLH85 resulted in significantly increased number and length of root hairs via ABA and auxin signaling pathways, increasing the absorption of Na+. Thus, SbbHLH85 plays a negative regulatory role in the salt tolerance of sorghum. We identified a potential interaction partner of SbbHLH85, which was phosphate transporter chaperone PHF1 and modulated the distribution of phosphate, through screening a yeast two-hybrid library. Both yeast two-hybrid and BiFC experiments confirmed the interaction between SbbHLH85 and PHF1. The overexpression of SbbHLH85 led to a decrease in the expression of PHF1 as well as the content of Pi. Based on these results, we suggested that the increase in the Na+ content and the decrease in the Pi content resulted in the salt sensitivity of transgenic sorghum.


Introduction
The problem of soil salinization exists worldwide. Salt stress caused by soil salinization affects the growth, development, and harvest yield of plants; in serious cases, salt stress even leads to plant death (Asano et al. 2012;Sui et al. 2017). The most cost-effective way to use salinized soil is to develop salt-tolerant crop varieties based on the well-established knowledge of molecular mechanisms underlying plant salt resistance, thereby promoting agricultural production. Sweet sorghum (Sorghum bicolor (L.) Moench) is a crop with high sugar content and high biomass yield, which has the laudatory name of second-generation sugarcane (Sui et al. 2015). In today's world, sweet sorghum is an important food, feed, and energy crop. Besides high biomass, sweet sorghum also has tolerance to various abiotic stresses, especially to salt (Schnippenkoetter et al. 2017). Many genes have been found to play an important role in regulating salt tolerance in sweet sorghum (Song et al. 2020;Wang et al. 2014b;Yang et al. 2018;Zheng et al. 2011). It is believed that the outcomes from studying the characteristics of these genes can be applied to improve the salt tolerance in other crops, which also is of great significance for understanding plant growth and development in saline alkali environments.
Salt stress increases the level of sodium ion (Na + ) and potassium ion (K + ), thus reducing the level of nutrient elements (e.g., N and P) and causing the imbalance of nutrient distribution in plants (Colla et al. 2008;Wang et al. 2020). The phosphorus content in pomelo and orange decreases significantly in response to salt stress (Ma et al. 2005). Phosphorus (P) is an essential element required for plant growth Communicated by Emma Mace. Yushuang Song, Simin Li and Yi Sui authors have contributed equally to this work.

3
and development (Dalen 2012;Miao et al. 2009). Phosphorus transfer between different tissues and subcellular organelles in plants is facilitated by a series of transport proteins having phosphorus transfer activity (Deng et al. 2017). The amount of phosphotransporters on the plasma membrane directly determines the absorption and distribution efficiency of phosphorus in plants (Mudge et al. 2002). PHOSPHATE TRANSPORTER 1 (PHT1) encodes an inorganic phosphate (Pi) transporter, which is regulated by the plant-specific phosphate transport chaperone PHF1 during its transport to the plasma membrane and plays a fundamental role in the acquisition and reactivation of Pi (González et al. 2005;Shin et al. 2004).
The bHLH transcription factor proteins have a highly conserved basic/helix-loop-helix special domain, namely the bHLH domain Maia et al. 2012;Yao et al. 2018). bHLH transcription factors are involved in salt stress response (Chen et al. 2018(Chen et al. , 2017Zhou et al. 2009). The expression of both AtbHLH92 and OrbHLH001 was induced by salt, and their overexpression improved salt tolerance in plants (Jiang et al. 2009;Li et al. 2010). When the bHLH transcription factor AtNIG1 was knocked out, the resulting mutant became more sensitive to salt stress in Arabidopsis. AtNIG1 regulated salt tolerance by binding specifically to E-box sequences in the promoter regions of many salt stressrelated genes (Kim and Kim, 2006).
Studies showed that plant bHLH transcription factors were also involved in root hair formation (Gajewska et al. 2018;Yan et al. 2014). The root hair is a tubular protrusion formed by the extension of root-specific epidermal cells, which contains enzymes and nutrient transporters involved in nutrient absorption (Menand et al. 2007;Wei and Li 2018). Root hairs increase the contact area between plants and soil and determine the efficiency of root absorption of water and nutrients (Krasilnikoff et al. 2003). The Arabidopsis bHLH transcription factors GLABRA3 and ENHANCER OF GLABRA3 have redundant functions. The number of root hairs decreased slightly in the single mutants of these two genes, but decreased significantly in the double mutants (Bernhardt et al. 2005). The polar growth of root hairs is initially triggered by RHD6/RSL1 of the bHLH family, and then, its elongation is activated by the expression of RSL4/ RSL2 (Vijayakumar et al. 2016). RSL4 is essential for root hair elongation in Arabidopsis and controls the final root hair cell size (Zhu et al. 2020). RSL2 affects reactive oxygen species (ROS) production and root hair growth (Rymen et al. 2017).
In this study, we first cloned a bHLH gene, SbbHLH85, in sweet sorghum in M-81E, which was induced by salt stress and ABA (Liu et al. 2015a). Some reports showed the involvement of the bHLH transcription factor in salt stress response and root hair formation. We first overexpressed SbbHLH85 in sorghum and Arabidopsis to investigate the molecular mechanism underlying the regulation of root hair development by SbbHLH85 in response to salt stress in sweet sorghum. The results showed that, different from its homologs in other species, the sorghum SbbHLH85 was a new atypical bHLH transcription factor and a key gene for root development in sorghum. The overexpression of SbbHLH85 in sorghum and Arabidopsis significantly increased the number and length of root hairs. However, salt resistance was significantly lower in overexpression lines. The overexpression of SbbHLH85 can influence the expression of the genes involved in ABA and auxin signal transduction (PYL and PIN3), peroxidase (PERs), root hair development, and receptor-like proteins (RLKs). In addition, SbbHLH85 interacts directly with SbPHF1, a phosphate transporter chaperone protein in sorghum, affecting the transport of Pi. Based on these results, we suggested that SbbHLH85 participated in regulating ABA and auxin signal transduction pathways and distribution of nutrients to affect the development of root hairs, thus affecting the absorption of Na + and the content of ROS and mediating plant salt response.

Plant materials and growth conditions
In this study, wild-type Arabidopsis (WT) was used as a control. The WT and mutant seeds were evenly seeded on 1/2 Murashige and Skoog (MS) medium with corresponding antibiotics; the composition of 1/2 MS medium was as described previously . The culture dish was placed at low temperature for 3 days and cultured in the tissue culture room (22 °C-16-h light/18 °C-8-h dark), and the screened positive seedlings were transferred to nutrient soil for further cultivation until the seeds matured. The harvested seeds were seeded using the same method in 1/2 MS medium with salt, and the phenotype of the plants was observed.
Sweet sorghum cultivar M-81E was used in this experiment. The seeds of sweet sorghum were cultured in the sand. Tap water was irrigated before emergence, and 1/2 Hoagland nutrient solution was added every day after emergence. When the seedlings grew to the four-leaf stage, Hoagland nutrient solution containing 0, 50, 100, 150, and 200 mM NaCl was applied to sweet sorghum under salt stress.

Cloning, bioinformatics, and expression analysis of SbbHLH85
The SbbHLH85 CDS sequence was obtained by comparing the AtRSL2 sequence on the Ensembl website (http:// ensem bl. grame ne. org/). Online website NCBI (https:// www. ncbi.

3
nlm. nih. gov/), SMART (http:// smart. embl-heide lberg. de/), and software MEGA6 were used for nucleic acid and protein sequence analysis and the construction of evolutionary trees.
The root tissue of sweet sorghum was used for the cloning and expression of SbbHLH85. Samples were immediately frozen in liquid nitrogen and stored at -80 °C before analysis. Three biological replicates were used for qRT-PCR. Sweet sorghum internal reference gene Sbactin-1 was used as a control. The primers are listed in Supplementary  Table 1.

Subcellular localization of SbbHLH85 protein
The KpnI and BamHI sites were used to insert cloned SbbHLH85 CDS into pCAMBIA1300-35S-sGFP vectors to produce 35S: SbbHLH85-green fluorescent protein (GFP) constructs. These were transferred into Agrobacterium tumefaciens GV3101 and used to infect tobacco epidermal cells. GFP signal was observed using a two-photon laser scanning confocal microscope (TCS S8MP, Leica, Germany). 35S: GFP transgenic tobacco was used as a localization control for expression in the cytoplasm/nucleus. The primers are listed in Supplementary Table 1.

Generation of transgenic plants
For the overexpression of SbbHLH85 in Arabidopsis, SbbHLH85 genes were linked to pROKII vectors using the XbaI and KpnI restriction sites and transferred to A. tumefaciens GV3101 to obtain SbbHLH85-overexpressing plants by infecting WT inflorescence. The transgenic Arabidopsis plants were screened with kanamycin (50 g/mL) and verified using RT-PCR. The full-length cDNA of SbbHLH85 was inserted into the pMWB110 vector through BamHI and KpnI sites to obtain the pMWB110-SbbHLH85 vector. The pMWB110-SbbHLH85 vector was introduced into sorghum by Agrobacterium-mediated transformation. PCR, herbicide (glufosinate) treatment, and bar rapid detection kit were used to detect transgenic plants. The primers used are listed in Supplementary Table 1.

Agrobacterium-mediated genetic transformation in sorghum Tx430
This method was based on the transformation protocol in sorghum (Do et al. 2016), with some modifications. Briefly, the husks were removed 12-16 days after pollination, and the immature seeds were sterilized in 70% (v/v) ethanol for 5 min and then in 12% (v/v) bleach for 10-15 min. Finally, the seeds were rinsed three times with autoclaved water. Immature embryos, 1.0-1.5 mm in length, were isolated from immature seeds and placed in infection liquid medium (0.44 g /L Murashige-Skoog salts, 1 × B5 vitamins, 68 g/L sucrose, 36 g/L glucose, 1 g/L asparagine, 1 g/L casamino acids, 0.2 g/L cysteine, 2 mg/L 2,4-dichlorophenoxyacetic acid, and 200 μM acetosyringone, pH 5.2). The vector 85-OE and 85-CR were introduced into the Agrobacterium strain EHA105 by heat shock. Positive EHA105 cells were cultured in YEB medium (5 g/L beef extract, 5 g/L peptone, 1 g/L yeast extract, 5 g/L sucrose, and 10 mM magnesium sulfate, pH 7.0) overnight until optical density at 600 nm (OD 600 nm) was 1.0. About 100-200 immature embryos were subjected to heat treatment (43 °C for 3 min) before inoculation with 1 mL of bacterial cells for about 5 min. After infection, embryos were transferred to the co-cultivation medium (infection medium with 8 g/L agarose). The scutellum was faced up and co-cultivated at 25 °C in the dark for 2-3 days. After co-cultivation, the embryos were subcultured on CIM (Do et al. 2016) resting medium containing 250 mg/L timentin at 28 °C in the dark for 6-7 days and then transferred to the CIM selection medium containing 50 mg/L paromomycin for 85-CR or 5 mg/L bialaphos for 85-OE for about 10 days. Another 20-day selection was performed in the same medium until resistant calli were formed, and then transferred into SM medium (Do et al. 2016) (250 mg/L timentin, 50 mg/L paromomycin or 5 mg/L bialaphos, and 8 g/L agarose, pH 5.7) at 28 °C under 16-h light: 8-h dark conditions for 2-3 weeks. Elongated shoots, 1-3 cm in length, were transferred to the rooting medium [half-strength Murashige-Skoog basal salt with vitamins, 30 g/L sucrose, 0.1 g/L myo-inositol, and 2.6 g/L Gelzan (gellan gum), pH 5.6] for rooting. Putative transgenic plants with healthy roots were transferred into pots before moving to the field. The primers are listed in Supplementary Table 1.

Quantification of biomass, MDA content, and Na + and K + contents
For measuring biomass, we first took whole plants out of the pot, washed them, weighed them, and recorded the fresh weight. Then, the plants were dried in an oven for 7 days and then weighed again to record dry weight (105 °C, 15 min and 65 °C, 7 days). Fresh weight and dry weight were measured for each treatment using five replicates (Song et al. 2019). The MDA contents were determined as described by Ma (Ma et al. 2013): The leaves of 0.2 g of each line were weighed, and 5 mL of 0.1% trichloroacetic acid was added for grinding. The homogenate was mixed with 5 mL of 0.5% thiobarbituric acid (TBA), boiled for 10 min, taken out, cooled to room temperature, and centrifuged at 3000 rpm for 15 min. The supernatant was taken, and the volume was measured. The absorbance of the solution at wavelengths 532 nm and 600 nm was determined using an ultraviolet (UV) spectrophotometer. The blank control contained 0.5% TBA solution. The Na + and K + contents were determined by the specific steps described by Song: Each line was treated with 0 and 100 mM NaCl. After 10 days, 0.3 g roots were placed in 5 mL of ddH 2 O and boiled for 2 h. The plant residue was filtered, and the volume was made up to 10 mL. The contents of Na + and K + of each treated line were determined using a flame spectrophotometer (Song et al. 2020).

Diaminobenzidine and nitroblue tetrazolium staining
Arabidopsis seedlings were grown for about a month and treated with 0 or 100 mM NaCl in 1/2 concentration Hoagland solution for 48 h. The rosette leaves of WT and overexpressing plants with the same growth were put in diaminobenzidine (DAB) or nitroblue tetrazolium dye solution and placed in the dark for more than 12 h. They were then put in the bleach (3:1:1 ethanol: acetic acid: glycerol) and boiled in boiling water for 10-15 min for decolorization. The color change in the blade was observed, and images were taken.

Root hair experiment
The homologous genes of SbbHLH85 are AtRSL2 and AtRSL4 in Arabidopsis. AtRSL2 is closely related to root hair development and elongation, while AtRSL4 is a functional redundancy gene. Therefore, we selected WT, M-81E, overexpressing (At-OX4, At-OX13, Sb-OX1, Sb-OX3, Sb-OX6, and Sb-OX7), RSL2 mutant (rsl2-1 and rsl2-3), RSL4 mutant (rsl4), and RSL2 and RSL4 double mutant (rsl2rsl4) lines as experimental subjects. The main root hair development and the root hair elongation were tested. Each Arabidopsis seed was on demand in 1/2 MS medium. The sorghum seeds were hydroponic. The development and elongation of root hairs at the root tip 5 mm from the main root of each line were observed under an electron microscope after 7 days in the lab. The microscope magnified the images by 40 times.

RNA-seq assay
The roots of WT, overexpressing, mutant, and double mutant lines were collected and preserved in liquid nitrogen. RNAseq and differential gene expression analysis were carried out using llumina NovaSeq 6000 (Biomarker Technologies, Beijing, China). The transcriptome analysis of 24 samples was completed, and 156.37 Gb of clean data were obtained. The HISAT2 system was used to sequence the clean reads of each sample with the designated reference genome, and the reads was assembled by StringTie comparison.
The reads on the sample were assembled and quantified by StringTie comparison after the comparison and analysis. The gene expression was analyzed based on the comparison results. StringTie used fragments per kilobase of script per million fragments mapped as an index to measure the expression level of transcripts or genes. The differentially expressed genes were identified according to their expression levels in different samples, and Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Cluster of Orthologous Groups of proteins (KOG) databases were used for functional annotation and enrichment analysis. KEGG pathway analysis was carried out for the common differentially expressed genes in each comparison group. The heat map clustering analysis was performed for the expression level of selected genes, which was completed on the platform of BMKCloud.

RT-PCR validation of genes related to plant salt resistance and root hair development
We extracted RNA from the root tissues of WT, At-OX13, rsl2-3, and rsl2rsl4 lines, which had been growing for about 1 month, and reverse-transcribed it into cDNA, followed by RT-PCR. The primers used are listed in Supplementary  Table 2. In sorghum, M-81E and Sb-OX1 were treated with 0 mM and 100 mM NaCl for 48 h. Then, RNA was extracted and transformed into cDNA, following which RT-PCR was performed. The primers used are listed in Supplementary  Table 3.

Yeast two-hybrid experiment
The CDs of SbbHLH85 were cloned into the bait vector pGBKT7 to obtain SbbHLH85-BD and then transformed into yeast strain Y2HGold (Clontech). After detecting that the gene had no self-activating activity, we constructed a yeast two-hybrid library of sorghum, and hybridized and screened it according to the matching scheme described in Clontech's matchmaker gold yeast two-hybrid user's manual. After screening, the interaction between SbPHF1 and SbbHLH85 was verified by the yeast two-hybrid experiment. The full-length coding sequence (CDS) of SbPHF1 was cloned into pGADT7 to obtain SbPHF1-AD, and SbbHLH85-BD and SbPHF1-AD were co-transformed into Y2HGold. Then, the growth of the colony on the corresponding medium was observed. The primers used in the yeast two-hybrid experiment are listed in Supplementary  Table 1.

BiFC experiment
The CDs of SbbHLH85 and the N-terminal of pSPYNE-35S were fused to obtain SbbHLH85-N-YFP, while the CDs of SbPHF1, and the C-terminal of pSPYNE-35S were fused to obtain SbPHF1-C-YFP. The obtained plasmid was introduced into A. tumefaciens (GV3101), and transient transformation was used to infiltrate the tobacco. The fluorescence was observed under a confocal laser scanning microscope (Olympus) after 48 h of normal culture. The primers of BiFC are listed in Supplementary Table 1.

Measurement of the phosphorus (Pi) content
To measure the Pi content, we first washed the sweet sorghum leaf sheaths. The surface moisture was dried, and then, 0.3 g of the sheaths were weighed and ground in a mortar. Some quartz sand and 5 mL of distilled water were added; each treatment was performed in five replicates. The homogenate was transferred to a 25-mL volumetric flask. The residue was washed in the mortar and volumed to the scale; 10 mL of the supernatant of the mixed solution was centrifuged at 6000 rpm for 15 min. The absorbance of the solution at wavelength 660 nm was determined using a UV spectrophotometer. The measured absorbance value was substituted into the standard curve to calculate the Pi content.

Statistical analysis
The statistical results were described as mean ± standard deviation. SPSS ver. 17.0 statistical software was used to analyze the data. One-way analysis of variance was used as the designated package. Different letters indicated a significant difference in the average (0.05) of the Duncan test.

SbbHLH85 was a salt stress-responsive gene
For preliminary characterization of the function of SbbHLH85, we studied its conserved domain, expression pattern, and subcellular localization. First, the phylogenetic tree analysis based on the protein sequence showed that SbbHLH85 was closely related to A0A2S3HF40 in Panicum hallii and A0A3L6RBM2 in Panicum miliaceum. SbbHLH85 shared 33.15% and 32.18% homology with the Arabidopsis genes AtRSL2 and AtRSL4, respectively (Fig. 1a). After analyzing the domain structure, we found that all aforementioned proteins had a bHLH domain at the C end (Fig. 1b). The SbbHLH85 gene contained a 951-bp coding sequence that encoded 316 amino acids. Different from RSL2 and RSL4, the SbbHLH85 protein comprised a bHLH domain between amino acids 238 and 287, which contained a A4-R7 motif harboring an alanine at the fourth position and arginine at the seventh position ( Fig. 1c and Supplementary Fig. 1). The SbbHLH85 had a unique A4-R7 motif and hence could not directly bind to the promoter of the downstream genes (Toledo-Ortiz et al. 2003). The expression pattern of SbbHLH85 was further explored by measuring the relative abundance of SbbHLH85 in the roots of M-81E under different salt stresses. RT-PCR results showed that the expression of SbbHLH85 in sorghum roots decreased with the increase in the salt concentration, dropping to the lowest level at 100 mM NaCl (Fig. 1d). The subcellular distribution of SbbHLH85 was examined by fusing SbbHLH85 with GFP. In the lower epidermal cells of tobacco, signals were detected in the nucleus (Fig. 1e). Therefore, SbbHLH85 was a bHLH transcription factor negatively induced by salt and localized in the nucleus.

SbbHLH85 actively regulated the development of root hairs
We generated SbbHLH85-overexpressing lines At-OX4 and At-OX13 controlled by the CaMV 35S promoter in Arabidopsis to explore the effect of SbbHLH85 on root hair development (Supplementary Fig. 2). We used the AtRSL2 mutants rsl2-1 and rsl2-3 and the AtRSL4 mutant rsl4 to test the effect of SbbHLH85 on root development. All the Arabidopsis mutants were purchased from the TAIR website (https:// www. arabi dopsis. org/). The double mutant rsl2rsl4 was provided as a gift by Professor 1 3 Hongwei Guo of Southern University of Science and Technology. Subsequently, the complementing lines Crsl2-1, Crsl2-3, Crsl4, and Crsl2rsl4 were created by expressing SbbHLH85 in rsl2-1, rsl2-3, rsl4, and rsl2rsl4, respectively. Figure 2a shows that the number and length of root hairs in the overexpressing lines were the largest and the longest, whereas the control mutants rsl2-1, rsl2-3, and rsl4 had fewer and shorter root hairs. Further, no root hairs were observed in the double mutant rsl2rsl4. Thus, SbbHLH85 could complement the root hair defect in the single and double mutants (Fig. 2a-c).
We generated SbbHLH85-overexpressing lines controlled by the maize ubiquitin promoter in Tx430 (a sorghum inbred line) background and detected the relative expression of SbbHLH85 in each line by RT-PCR to further investigate the role of SbbHLH85 during salt stress in sorghum. Compared with the wild type, SbbHLH85 was highly expressed in Sb-OX1, Sb-OX3, Sb-OX6, and Sb-OX7 (Fig. 2d). It was also found that SbbHLH85 overexpression increased the number and length of root hairs in sorghum ( Fig. 2e-g).

SbbHLH85 negatively regulated the salt stress response
We tested each of the Arabidopsis overexpressing lines to explore the role of SbbHLH85 in regulating salt stress response in plants. We studied the effects of salt stress on plant growth, including germination rate and seedling survival. For the germinating plants, we sowed seeds of each line in 1/2 MS medium containing 0, 100, or 150 mM NaCl. The plants with overexpression grew more slowly than WT and the mutant lines in the salt medium (Fig. 3a). The germination rate and root length of plants with overexpression were higher than those in WT and mutants irrespective of the nonstressed or salt stress condition (Fig. 3b  and c). Under abiotic stress, plants were exposed to ion stress, oxidative stress, and osmotic stress at the same time. Next, we compared the germination rate and root length of each line under different stress conditions. The growth of seedlings in NaCl, LiCl, and mannitol media showed different degrees of slow growth ( Supplementary  Fig. 3a). Among these, NaCl treatment resulted in the worst growth, the shortest taproot length, and the lowest germination rate (Supplementary Fig. 3b and 3c). We also studied the salt tolerance of different lines in the seedling stage. Salt stress reduced the biomass of each line, but the fresh weight and dry weight of mutant lines decreased to a lesser extent ( Fig. 3d and Supplementary Fig. 4a). The contents of MDA increased more in WT plants and plants with overexpression, but less in mutant lines (Supplementary Fig. 4b). The ion content analysis under salt stress showed that the Na + content in overexpressing lines were higher than that in WT control (Fig. 3e), while the trend was opposite for the K + content (Fig. 3f) Fig. 3d and e). We also transformed SbbHLH85 into the mutant lines to investigate whether SbbHLH85 could rescue the salt-sensitive phenotype. The results showed that the physiological indexes of the complemental lines Crsl2-1, Crsl2-3, Crsl4, and Crsl2rsl4 under salt stress recovered ( Fig. 4; Supplementary Fig. 4c and d). Overall, these findings suggested that the ectopic overexpression of SbbHLH85 in Arabidopsis might affect the homeostasis of Na + and K + , the content of ROS, and the degree of membrane lipid peroxidation to improve salt tolerance.
We further tested the salt tolerance of the SbbHLH85overexpressing lines in sorghum to better understand the function of SbbHLH85 in salt stress. Under salt stress, the overexpressing lines in sorghum showed obvious weak growth, yellow leaves, and even curly symptoms (Fig. 5a). Next, we studied the biological processes of salt stress. The biomass accumulation in overexpressing lines was lower than that in M-81E (Fig. 5b and c). Under salt stress, the content of Na + in overexpressing lines increased significantly and the content of K + decreased ( Fig. 5d and e; Supplementary Fig. 5a). Compared with the WT of M-81E, MDA content of overexpressing lines increased significantly after salt treatment (Supplementary Fig. 5b). This was consistent with the conclusion from Arabidopsis that SbbHLH85 participated in salt stress response.

Transcriptome analysis of plants with an altered SbbHLH85 level
We analyzed the RNA-seq data of WT, overexpressing, mutant, and double mutant lines under control and NaCl treatments to reveal the molecular mechanism of SbbHLH85 regulating plant response to salt stress and root hair development. We selected 22 representative genes for RT-PCR verification to verify the accuracy of transcriptome, and found that the transcriptome data were of high quality ( Supplementary Fig. 6b). In the RNA-seq experiment, we used 48-h NaCl treatment and three biological replicates. Under salt stress, 597 differentially expressed genes (DEGs) were found in WT, 244 DEGs in overexpressing lines, 204 DEGs in single mutants, and 278 DGEs in double mutants. We intersected these DGEs and got 156 DGEs (Supplementary Fig. 6a). Then, we used hierarchical clustering and correlation analysis to analyze the expression patterns of 156 DEGs. The GO analysis showed that these DEGs were involved in biological process, molecular function, and cellular component; of these, they are mainly involved in biological process ( Supplementary Fig. 7). The KEGG and KOG analyses showed that these DEGs were mainly related to phenylalanine metabolism, hormone signal transduction, and secondary metabolite metabolism ( Supplementary Fig. 6c and 8). After further analysis of the genes in these pathways, we found that they were mainly involved in auxin signaling pathway (AtPIN3 and AtSAUR50), ABA signaling pathway (AtPYL6), and also the production of peroxidase (PER) (AtPRX33, AtPRX37, AtRCI3, and AT4G08780) and receptor-like kinases (RLK) (AT4G00970 and AT4G04570) (Fig. 6a) and development of root hairs (Fig. 6b). This was consistent with our previous conclusion that SbbHLH85 participated in salt stress and affected the development of root hair.
SbbHLH85 had a unique A4-R7 motif and hence could not directly bind to the promoter of the downstream genes. We identified all sorghum genes homologous to the Arabidopsis ones to further study the molecular mechanism of SbbHLH85 regulating salt response and root hair development of sorghum. These genes were annotated as auxin signal, ABA signal, root hair development, PER, and RLK. This study identified one ABA signal transduction gene (SbPYL4), one auxin signal transduction gene (SbPIN3), two root hair development-related genes (SbRSH2 and SbRHL1), four PER genes (SbPER3, SbGLO1, SbPER4, and SbPER35), and three RLK-related genes (SbRLK1, SbRLK2, and SbRLK8) in sorghum (Fig. 6c). This was consistent with the previous findings, indicating that SbbHLH85 was involved in root hair development and salt tolerance of sorghum.

SbbHLH85 interacted with SbPHF1 and regulated Pi accumulation to participate in salt tolerance
Increasing lines of evidence showed that the bHLH proteins acted by forming protein complexes with other interacting proteins (Abe et al. 2003;Oh et al. 2007). We used a yeast two-hybrid system to screen for interactors so as to find the potential chaperone of the bHLH85 protein. First, the BD domain of pGBKT7 and bHLH85 was fused as the bait. After, we proved that bHLH85 had no self-activating activity (Fig. 7a); the cDNA library containing the prey protein insert fused with GAL4-AD was used to co-transform the yeast cells with SbbHLH85-BD. Three colonies were positive for X-a-gal and ABA. Among these candidates, only the binding of SbPHF1 and SbbHLH85 was stable. SbPHF1 encodes a SEC12-like protein and is homologous to AtPHF1. As a chaperone of phosphate transporter PHT1, SbPHF1 helps transport PHT1 to the plasma membrane. The SbPHF1-AD vector was co-transformed into the Y2H competent state with the SbbHLH85-BD vector to confirm their interaction in yeast. As shown in Fig. 7b, the lines in the experimental group grew normally on the medium, which was mixed with X-a-gal but lacked isoleucine and tryptophan (SD/-L/-T/X), and the colony turned blue. Similarly, the lines in the experimental group grew normally on the medium mixed with X-agal and ABA but lacking isoleucine, tryptophan, histidine, and adenine (SD/-L/-T/-H/-Ade/X/A), and the colony turned blue. This finding showed that SbbHLH85 had a strong interaction with SbPHF1 (Fig. 7b).
We then used a BiFC system to verify the aforementioned observation so as to determine whether the interactions also existed in plant cells. The A. tumefaciens lines with SbPHF1-C-YFP and SbbHLH85-N-YFP were mixed and transfected into tobacco leaves. At the same time, the empty carrier was combined with each fusion structure and injected into tobacco leaves. After incubation for 2 days, a YFP signal was observed under a fluorescence microscope. The co-transformed samples showed YFP fluorescence in the nucleus, while none of the control samples showed any YFP signal (Fig. 7c). This demonstrated that SbbHLH85 and SbPHF1 were co-localized and interacted in the plant nucleus.
We examined the expression of PHF1 and PHT1 and the content of Pi in SbbHLH85-overexpressing and WT sorghum to investigate the effect of the interaction between SbbHLH85 and SbPHF1. The results showed that the overexpression of SbbHLH85 led to a decrease in the expression of PHF1 and PHT1 as well as the content of Pi (Fig. 7d-f).

Discussion
How plants find a balance between environmental stress and plant growth is a new and important research topic (Monlau et al. 2015). Many indirect findings have confirmed that the bHLH proteins are involved in salt tolerance in plants (Babitha et al. 2015;Long et al. 2010;Waseem et al. 2019). The overexpression of SbbHLH85 is proved to increase the salt sensitivity of sorghum. The possible mechanism is that SbbHLH85 and SbPHF1 interaction affects the development of root hair through ABA and auxin signal transduction pathways, and hence the distribution of nutrients in plants, thus regulating the salt tolerance of sorghum. This observation enriched the regulation network of the salt stress response, and might be of great significance for improving crop productivity under adverse environmental conditions.

SbbHLH85, different from its homologs in other species, was a new atypical bHLH transcription factor
The bHLH transcription factors belong to the second-largest family of transcription factors in plants, named so for its bHLH domain (Herbst and Kolligs 2008). The family plays a key role in plant growth and abiotic stress (Jiang et al. 2009). In animals, bHLH transcription factors can be divided into six categories: A-F (Wang et al. 2010). The most bHLH transcription factors in plants belong to class B. Only 11% of plant bHLH proteins have a conserved motif: A4-R7, which is not found in animals (Sailsbery and Dean 2012). With the development of molecular biology, an increasing number of bHLH transcription factors have been found, especially the identification of new atypical bHLH transcription factors, which makes this family more diverse. We analyzed the protein sequence of SbbHLH85 and bHLH transcription factors in other species and found that RSL2 and RSL4 were the most closely related genes in Arabidopsis. However, different from RSL2 and RSL4, SbbHLH85 had an A4-R7 motif as shown in Supplementary Fig. 1, indicating that SbbHLH85 was a new atypical bHLH transcription factor. On the one hand, it enriched SbbHLH85 interacted with SbPHF1 and regulated phosphorus (Pi) accumulation to participate in salt tolerance. a Self-activation activity of SbbHLH85 was verified. SbbHLH85 was used as the bait protein. The self -activation activity of SbbHLH85 was verified by the yeast system on the medium lacking tryptophan (SD/-Trp), medium mixed with X-a-gal but lacking tryptophan (SD/-Trp/X), and medium mixed with X-a-gal and AbA but lacking tryptophan (SD/-Trp/X/A). b Yeast two-hybrid assays. Interaction was indicated by the ability of cells to grow in SD/-L/-T/X and SD/-L/-T/-H/-Ade/X/A medium. The Gal4 DNA -binding domain was fused with SbbHLH85 (shown as SbbHLH85-BD), while the Gal4 activation domain was fused with SbPHF1 (shown as SbPHF1-AD). c BiFC analysis. Fluorescence was observed in the nuclear chamber of lower epidermal cells of tobacco leaves. The C-terminal part of YFP was fused with SbPHF1 (SbPHF1-C-YFP), while the N-terminal part of YFP was fused with SbbHLH85 (SbbHLH85-N-YFP). No signals were observed from the negative controls. DAPI, 4',6-Diamidino-2-phenylindole. d and e Expression of SbPHF1 and PHT1 in WT and overexpressing lines of sorghum under salt stress. f Pi contents of WT and overexpressing lines of sorghum under salt stress. Data are presented as the mean ± SD of five measurements. Means with different letters are significantly different at P < 0.05 1 3 the diversity of the bHLH family in sorghum. On the other hand, it indicated that SbbHLH85 might play a unique role in promoting salt tolerance and growth of sorghum, which might be of great significance for crop improvement.

SbbHLH85 affected plant salt tolerance and root hair growth through ABA and auxin signal transduction pathways
ABA and auxin signal transduction pathways play an important role in regulating plant salt tolerance (Huang et al. 2012;Min et al. 2015;Sun et al. 2016) and root hair growth (Guo et al. 2018(Guo et al. , 2019Sun et al. 2019; Van et al. 2017). For example, PYR/PYL act as ABA receptors and bind to PP2C family members represented by AB12 and ABI1 to regulate the phosphorylation of downstream protein kinases and thus initiate the ABA signal transduction pathway to respond to salt stress (Aleman et al. 2016;Fujii et al. 2009;Nishimura et al. 2009;Raghavendra et al. 2010;Umezawa et al. 2009;Wang et al. 2014aWang et al. , 2017. The PINFORMED (PIN) protein family in the growth hormone regulation pathway is involved in regulating root growth and salt tolerance (Ganguly et al. 2012;Harrison and Masson 2008;Lewis et al. 2011;Liu et al. 2015b;Lv et al. 2018;Schlicht et al. 2008;Sun et al. 2008). In this study, the expression of ABA and auxin pathway genes, peroxidase, receptor-like protein kinase, and root hair development-related genes were changed by SbbHLH85 overexpression (Fig. 6a-c).
The homolog RSL4/RSL2 of SbbHLH85 in Arabidopsis is related to the development of root hairs; its expression is also regulated by exogenous hormones and environmental changes (Guo et al. 2018(Guo et al. , 2019Menand et al. 2007). Auxin can activate RSL4 expression and control ROS-related genes, including three peroxidases (PER1, PER44, and PER73) (Mangano et al. 2017). Similarly, RSL2 can inhibit the growth-promoting effect of auxin by inhibiting ROS produced by peroxides (PERs) (Kwon et al. 2015;Mangano et al. 2017;Wanapu and Shinmyo 1996). In this study, both Arabidopsis and sorghum overexpressing SbbHLH85 showed a decrease in the ROS content and a phenotype of root hair elongation.

SbbHLH85 negatively regulated plant salt resistance by affecting plant nutrient distribution
Under salt stress, the uptake of K + by root cells decreases and the absorption of Na + increases, leading to the imbalance of ion homeostasis (Song et al. 2020). Recent studies have shown that plants can protect themselves by reducing the length and density of root hair and the absorption area of excessive Na + when they sense stress signals (Lv et al. 2018;Wang et al. 2008).
Salt stress can also cause an imbalance of nutrient distribution in plants, including phosphorus (P) (Yang et al. 2003). P is a key element of many biomolecules (nucleic acids, ATP, and phospholipids) in many metabolic pathways; it is one of the important nutrients needed for plant growth and development (Marschner et al. 2005). Phosphate transporters encoded by the phosphate transporter family (PHT1) genes are important proteins for plant acquisition and transport of phosphates (Shin et al. 2004). PHT1;5 is strongly induced in the root, leading to the changes in Pi mobilization between plant roots/grounds. Compared with WT, PHT1;5 overexpression enhances root hair formation (Nagarajan 2010).
The transport of the PHT1 protein to the plasma membrane is regulated by plant-specific companion protein PHF1. The PHF1 gene encodes a plant-specific protein related to the SEC12 protein structure, which is located in the endoplasmic reticulum and is strongly expressed in plant root cuticle, root hair, and its cortical cells (Bayle et al. 2011). In 2005, a mutant of PHF1 was isolated in Arabidopsis. Studies on the mutant showed that the mutation of PHF1 reduced the accumulation of PHT1;1 transporter in the plasma membrane, destroyed the transport of Pi, and reduced the accumulation of Pi. In this study, SbbHLH85 interacted strongly with SbPHF1. Our result suggested that the interaction between SbbHLH85 and SbPHF1 might destroy the transportation and accumulation of Pi in plants under salt stress and aggravate the In conclusion, SbbHLH85, different from RSL2 and RSL4 in Arabidopsis, is a necessary gene for root development in sorghum. We then showed the molecular mechanism of SbbHLH85 regulating the salt response of sorghum by directly interacting with the specific companion protein PHF1, which can affect the distribution of Pi. SbbHLH85 participates in the ABA and auxin pathway mediated root hair development by affecting the expression of ABA and auxin pathway genes. The increase in the number and length of root hairs can promote the absorption of Na + . The increased Na + absorption and the decreased Pi content can ultimately result in the salt-sensitive phenotype of sorghum (Fig. 8). In future, it will be an effective measure to improve the salt tolerance of crops by properly reducing the number and length of root hairs in the presence of stress.